| Message ID | 20241021051151.4664-1-suhua.tanke@gmail.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | memblock: uniformly initialize all reserved pages to MIGRATE_MOVABLE | expand |
From: Mike Rapoport (Microsoft) <rppt@kernel.org> On Mon, 21 Oct 2024 13:11:51 +0800, Hua Su wrote: > Currently when CONFIG_DEFERRED_STRUCT_PAGE_INIT is not set, the reserved > pages are initialized to MIGRATE_MOVABLE by default in memmap_init. > > Reserved memory mainly store the metadata of struct page. When > HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=Y and hugepages are allocated, > the HVO will remap the vmemmap virtual address range to the page which > vmemmap_reuse is mapped to. The pages previously mapping the range will > be freed to the buddy system. > > [...] Applied to for-next branch of memblock.git tree, thanks! [1/1] memblock: uniformly initialize all reserved pages to MIGRATE_MOVABLE commit: ad48825232a91a382f665bb7c3bf0044027791d4 tree: https://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock branch: for-next -- Sincerely yours, Mike.
diff --git a/mm/mm_init.c b/mm/mm_init.c index 4ba5607aaf19..6dbf2df23eee 100644 --- a/mm/mm_init.c +++ b/mm/mm_init.c @@ -722,6 +722,10 @@ static void __meminit init_reserved_page(unsigned long pfn, int nid) if (zone_spans_pfn(zone, pfn)) break; } + + if (pageblock_aligned(pfn)) + set_pageblock_migratetype(pfn_to_page(pfn), MIGRATE_MOVABLE); + __init_single_page(pfn_to_page(pfn), pfn, zid, nid); } #else
Currently when CONFIG_DEFERRED_STRUCT_PAGE_INIT is not set, the reserved pages are initialized to MIGRATE_MOVABLE by default in memmap_init. Reserved memory mainly store the metadata of struct page. When HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=Y and hugepages are allocated, the HVO will remap the vmemmap virtual address range to the page which vmemmap_reuse is mapped to. The pages previously mapping the range will be freed to the buddy system. Before this patch: when CONFIG_DEFERRED_STRUCT_PAGE_INIT is not set, the freed memory was placed on the Movable list; When CONFIG_DEFERRED_STRUCT_PAGE_INIT=Y, the freed memory was placed on the Unmovable list. After this patch, the freed memory is placed on the Movable list regardless of whether CONFIG_DEFERRED_STRUCT_PAGE_INIT is set. Eg: Tested on a virtual machine(1000GB): Intel(R) Xeon(R) Platinum 8358P CPU After vm start: echo 500000 > /proc/sys/vm/nr_hugepages cat /proc/meminfo | grep -i huge HugePages_Total: 500000 HugePages_Free: 500000 HugePages_Rsvd: 0 HugePages_Surp: 0 Hugepagesize: 2048 kB Hugetlb: 1024000000 kB cat /proc/pagetypeinfo before： Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 … Node 0, zone Normal, type Unmovable 51 2 1 28 53 35 35 43 40 69 3852 Node 0, zone Normal, type Movable 6485 4610 666 202 200 185 208 87 54 2 240 Node 0, zone Normal, type Reclaimable 2 2 1 23 13 1 2 1 0 1 0 Node 0, zone Normal, type HighAtomic 0 0 0 0 0 0 0 0 0 0 0 Node 0, zone Normal, type Isolate 0 0 0 0 0 0 0 0 0 0 0 Unmovable ≈ 15GB after： Free pages count per migrate type at order 0 1 2 3 4 5 6 7 8 9 10 … Node 0, zone Normal, type Unmovable 0 1 1 0 0 0 0 1 1 1 0 Node 0, zone Normal, type Movable 1563 4107 1119 189 256 368 286 132 109 4 3841 Node 0, zone Normal, type Reclaimable 2 2 1 23 13 1 2 1 0 1 0 Node 0, zone Normal, type HighAtomic 0 0 0 0 0 0 0 0 0 0 0 Node 0, zone Normal, type Isolate 0 0 0 0 0 0 0 0 0 0 0 Signed-off-by: Hua Su <suhua.tanke@gmail.com> --- mm/mm_init.c | 4 ++++ 1 file changed, 4 insertions(+)